The absence of structural fidelity precludes construct validity: Rejoinderto Naglieri on what the cognitive assessment system does and does not measure
Tz. Keith et Jh. Kranzler, The absence of structural fidelity precludes construct validity: Rejoinderto Naglieri on what the cognitive assessment system does and does not measure, SCH PSYCH R, 28(2), 1999, pp. 303-321
In this rejoinder, we show that Naglieri's (1999a) commentary thoroughly fa
ils to refute the results of Kranzler and Keith's (1999) research on the fa
ctor structure of the Cognitive Assessment System (CAS). We begin by review
ing contemporary theory of construct validity with an emphasis on the role
of factor analysis in "strong" validation programs. Next, we respond to the
main theoretical and empirical issues raised by Naglieri (1999a) in his co
mmentary regarding the results of our confirmatory factor analyses of the s
tandardization data of the GAS. Finally, we critique the external evidence
of validity he presented and emphasize the need for further research on the
factor structure of the GAS. We also express concern with Naglieri's conti
nued selective and misleading use of quotations. In sum, our research indic
ates that the CAS lacks structural fidelity, a necessary but not sufficient
condition for construct validity. An absence of structural fidelity indica
tes that the scaled scores derived from the CAS do not correspond to the un
derlying theory upon which the: test is based. The external evidence of val
idity presented by Naglieri in his rejoinder is therefore meaningless. Scho
ol psychologists and other practitioners are advised to interpret the CAS w
ith caution, if at all.